Comparison of semistructured interviews with a quantitative measure of moral orientation

Journal of College Student Development, Mar/Apr 1998 by Liddell, Debora L

The MMO consists of two components: the first component presents nine first-person stories, followed by reaction statements to which the student responds using a Likert-type scale. Dilemmas and reaction statements were developed from interviews with students about their lives and relationships (Liddell et al., 1992). These reaction statements make up the care and justice scales. The second component of the MMO is a 14-item self-description inventory designed to measure students' perceptions of themselves as just, caring, or both. The MMO takes approximately 25 minutes to complete.

Personal Data Questionnaire. Participants completed a brief questionnaire identifying age, gender, ethnicity, year in school, and intended major.

Semistructured Interviews. Interviews were conducted by the author and a doctoral student, following the protocol established by Gilligan (1982) and Lyons (1988) designed to measure moral voice (see Appendix A), and modified for this study. The interview took approximately 30 minutes, and was tape-recorded. Prior to coding, the interview was transcribed and all identifying information such as gender and ethnicity was removed from the transcript.

The interview consisted of three components (Appendix A): Self-description (e.g., "How would you describe yourself to yourself?"), moral dilemmas ("Can you describe a situation you've been in when you weren't sure what was the right thing to do?"), and a discussion of the moral issues underlying one of the dilemmas that appeared in the MMO (Appendix B). The data were analyzed for the participants' descriptions of self, then for the considerations of moral voice in the context of real-life dilemmas, and finally for the considerations of Care or Justice in response to the specific MMO-generated dilemma. The MMO dilemma that was used for discussion stimulus was about plagiarism. Interviewees were asked to read the dilemma and were then asked a series of questions about the problem.

Two researchers independently rated the transcribed interviews, following Lyons' (1988) protocol for scoring. When ratings were completed, the two raters discussed any disagreements in their ratings, and determined compromise ratings. Scoring was determined by breaking down the transcribed interviews according to "considerations," or responses. Considerations were categorized as either response (Care) or rights (Justice) and scored by counting the number of considerations each individual presented within either mode. The predominance of Justice and Care orientations was determined by summing those considerations. Low or high scores can occur on either or both orientations; that is, Care and Justice are neither mutually exclusive, nor dichotomously scored.

Correlational analyses were used to answer the first three research questions. To answer the final research question, t tests were conducted. The grouping variable used in the t tests was moral voice. This grouping was accomplished by categorizing the interview scores using a median split, creating a categorical variable (high-low) for both moral orientations (Care and Justice) as well as the self-descriptions (Connected and Separate). In other words, interviewees with Care scores above the group median were assigned to the High-Care group; those with Care scores below the median were assigned to the Low-Care group. The same assignment decisions were made using scores from the Justice, Separate, and Connected constructs.